Persistent Reservation effect on non-medium commands

Gerry_Houlder at notes.seagate.com Gerry_Houlder at notes.seagate.com
Wed Dec 24 13:55:34 PST 1997


* From the T10 (formerly SCSI) Reflector (t10 at symbios.com), posted by:
* Gerry_Houlder at notes.seagate.com
*
The SPC (and other) standards are unclear what the response should be for
Persistent Reservations on commands that don't access the medium. I,m sure
this has been discussed but the result certainly isn't well documented.

Take the TEST UNIT READY command, for example. A regular logical unit
reservation (using RESERVE(6) command) requires this command to end with
RESERVATION CONFLICT status. This is appropriate because both READ and
WRITE commands will end with RESERVATION CONFLICT.

But what is the required response for a Persistent Reservation of type 1
(writes exclusive, reads shared) scope 0 (entire logical unit)? If either a
READ or a WRITE is allowed (but not both), should the TEST UNIT READY
command end with GOOD status? This is more in keeping with the wording for
extent reservations (which only apply to a subset of the logical blocks)
but it isn't obvious that this behavior is intended to apply when reads are
allowed but writes are prohibited.

I wish a general rule of thumb could be added to the model wording to
clarify this. What should be said about this?


*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at symbios.com




More information about the T10 mailing list